[FFmpeg-cvslog] avutil/{avstring,bprint}: add XML escaping from ffprobe to avutil

[Stefano Sabatini]

ffmpeg | branch: master | Stefano Sabatini <stefasab at gmail.com> | Mon Nov 16 09:38:39 2020 +0200| [0f6bf94eb71c2d5e996c89c290f1a53660c46c2e] | committer: Jan Ekström

avutil/{avstring,bprint}: add XML escaping from ffprobe to avutil

Base escaping only escapes values required for base character data
according to part 2.4 of XML, and if additional flags are added
single and double quotes can additionally be escaped in order
to handle single and double quoted attributes.

Co-authored-by: Jan Ekström <jan.ekstrom at 24i.com>
Signed-off-by: Jan Ekström <jan.ekstrom at 24i.com>

> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=0f6bf94eb71c2d5e996c89c290f1a53660c46c2e
---

 libavutil/avstring.h | 14 ++++++++++++++
 libavutil/bprint.c   | 29 +++++++++++++++++++++++++++++
 libavutil/version.h  |  2 +-
 tools/ffescape.c     |  7 +++++--
 4 files changed, 49 insertions(+), 3 deletions(-)

diff --git a/libavutil/avstring.h b/libavutil/avstring.h
index ee225585b3..fae446c302 100644
--- a/libavutil/avstring.h
+++ b/libavutil/avstring.h
@@ -324,6 +324,7 @@ enum AVEscapeMode {
     AV_ESCAPE_MODE_AUTO,      ///< Use auto-selected escaping mode.
     AV_ESCAPE_MODE_BACKSLASH, ///< Use backslash escaping.
     AV_ESCAPE_MODE_QUOTE,     ///< Use single-quote escaping.
+    AV_ESCAPE_MODE_XML,       ///< Use XML non-markup character data escaping.
 };
 
 /**
@@ -343,6 +344,19 @@ enum AVEscapeMode {
  */
 #define AV_ESCAPE_FLAG_STRICT (1 << 1)
 
+/**
+ * Within AV_ESCAPE_MODE_XML, additionally escape single quotes for single
+ * quoted attributes.
+ */
+#define AV_ESCAPE_FLAG_XML_SINGLE_QUOTES (1 << 2)
+
+/**
+ * Within AV_ESCAPE_MODE_XML, additionally escape double quotes for double
+ * quoted attributes.
+ */
+#define AV_ESCAPE_FLAG_XML_DOUBLE_QUOTES (1 << 3)
+
+
 /**
  * Escape string in src, and put the escaped string in an allocated
  * string in *dst, which must be freed with av_free().
diff --git a/libavutil/bprint.c b/libavutil/bprint.c
index 2f059c5ba6..e12fb263fe 100644
--- a/libavutil/bprint.c
+++ b/libavutil/bprint.c
@@ -283,6 +283,35 @@ void av_bprint_escape(AVBPrint *dstbuf, const char *src, const char *special_cha
         av_bprint_chars(dstbuf, '\'', 1);
         break;
 
+    case AV_ESCAPE_MODE_XML:
+        /* escape XML non-markup character data as per 2.4 by default: */
+        /*  [^<&]* - ([^<&]* ']]>' [^<&]*) */
+
+        /* additionally, given one of the AV_ESCAPE_FLAG_XML_* flags, */
+        /* escape those specific characters as required. */
+        for (; *src; src++) {
+            switch (*src) {
+            case '&' : av_bprintf(dstbuf, "%s", "&");  break;
+            case '<' : av_bprintf(dstbuf, "%s", "<");   break;
+            case '>' : av_bprintf(dstbuf, "%s", ">");   break;
+            case '\'':
+                if (!(flags & AV_ESCAPE_FLAG_XML_SINGLE_QUOTES))
+                    goto XML_DEFAULT_HANDLING;
+
+                av_bprintf(dstbuf, "%s", "'");
+                break;
+            case '"' :
+                if (!(flags & AV_ESCAPE_FLAG_XML_DOUBLE_QUOTES))
+                    goto XML_DEFAULT_HANDLING;
+
+                av_bprintf(dstbuf, "%s", """);
+                break;
+XML_DEFAULT_HANDLING:
+            default: av_bprint_chars(dstbuf, *src, 1);
+            }
+        }
+        break;
+
     /* case AV_ESCAPE_MODE_BACKSLASH or unknown mode */
     default:
         /* \-escape characters */
diff --git a/libavutil/version.h b/libavutil/version.h
index b7c5892a37..356c54d633 100644
--- a/libavutil/version.h
+++ b/libavutil/version.h
@@ -79,7 +79,7 @@
  */
 
 #define LIBAVUTIL_VERSION_MAJOR  56
-#define LIBAVUTIL_VERSION_MINOR  66
+#define LIBAVUTIL_VERSION_MINOR  67
 #define LIBAVUTIL_VERSION_MICRO 100
 
 #define LIBAVUTIL_VERSION_INT   AV_VERSION_INT(LIBAVUTIL_VERSION_MAJOR, \
diff --git a/tools/ffescape.c b/tools/ffescape.c
index 0530d28c6d..1ed8daa801 100644
--- a/tools/ffescape.c
+++ b/tools/ffescape.c
@@ -78,8 +78,10 @@ int main(int argc, char **argv)
             infilename = optarg;
             break;
         case 'f':
-            if      (!strcmp(optarg, "whitespace")) escape_flags |= AV_ESCAPE_FLAG_WHITESPACE;
-            else if (!strcmp(optarg, "strict"))     escape_flags |= AV_ESCAPE_FLAG_STRICT;
+            if      (!strcmp(optarg, "whitespace"))        escape_flags |= AV_ESCAPE_FLAG_WHITESPACE;
+            else if (!strcmp(optarg, "strict"))            escape_flags |= AV_ESCAPE_FLAG_STRICT;
+            else if (!strcmp(optarg, "xml_single_quotes")) escape_flags |= AV_ESCAPE_FLAG_XML_SINGLE_QUOTES;
+            else if (!strcmp(optarg, "xml_double_quotes")) escape_flags |= AV_ESCAPE_FLAG_XML_DOUBLE_QUOTES;
             else {
                 av_log(NULL, AV_LOG_ERROR,
                        "Invalid value '%s' for option -f, "
@@ -104,6 +106,7 @@ int main(int argc, char **argv)
             if      (!strcmp(optarg, "auto"))      escape_mode = AV_ESCAPE_MODE_AUTO;
             else if (!strcmp(optarg, "backslash")) escape_mode = AV_ESCAPE_MODE_BACKSLASH;
             else if (!strcmp(optarg, "quote"))     escape_mode = AV_ESCAPE_MODE_QUOTE;
+            else if (!strcmp(optarg, "xml"))       escape_mode = AV_ESCAPE_MODE_XML;
             else {
                 av_log(NULL, AV_LOG_ERROR,
                        "Invalid value '%s' for option -m, "